See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/27526767

Implementation of Engine Loss Analysis Methods in the Numerical Propulsion

System Simulation

Article · January 2005

DOI: 10.1115/GT2005-68202 · Source: OAI

CITATIONS READS
2 68

3 authors, including:

Travis Danner
gold coast inc.
3 PUBLICATIONS   7 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Travis Danner on 16 August 2017.

The user has requested enhancement of the downloaded file.

 Proceedings of ASME Turbo Expo 2005:
Land Sea & Air
June 6-9, 2005, Reno-Tahoe, Nevada, USA

GT2005-68202

IMPLEMENTATION OF ENGINE LOSS ANALYSIS METHODS IN THE NUMERICAL

PROPULSION SYSTEM SIMULATION

Bryce A. Roth, Erin M. McClure, Travis W. Danner

Georgia Institute of Technology
Atlanta, GA 30332-0150

ABSTRACT analysis [1,2]. Past cycle codes were not readily extensible or
This paper describes the implementation and application of flexible enough to easily accommodate post-facto incorporation
a new set of thermodynamic loss analysis tools in the of loss calculation routines. As a result, the cycle analyst is
Numerical Propulsion System Simulation. This analysis tool usually forced to do this type of analysis by hand in a
set is intended to enable fast, accurate estimation of losses in an spreadsheet environment for a limited set of engine operating
engine cycle model with minimal effort on the part of the user. conditions (if it is performed at all).
The basic thermodynamic concepts and analysis methods are The advent of modern cycle analysis codes implemented in
first described. Next, the implementation of the necessary object-oriented languages has largely removed this barrier.
thermodynamic calculation functions is described. These Such cycle analysis codes offer the ability to seamlessly add
functions are intended to be used in conjunction with a general- objects and functions required for loss analysis without the
purpose loss analysis element to facilitate estimation of all need for major modification of the existing code base. The
losses in an engine cycle model. The loss analysis element is research effort described in this paper represents a first step
described in detail and is subsequently used to analyze a mixed towards the construction of a suite of general tools to facilitate
flow turbofan engine. Typical performance and loss results are loss analysis in aircraft engine cycle models.
presented. The resultant detailed loss information is not This paper begins with a discussion on the fundamental
normally available when using standard cycle analysis concepts and definitions of loss and work potential and how
methods. The information gained from this analysis is useful in they can be applied to the analysis of engine performance.
that it yields insight into the underlying losses that contribute to Next, the implementation of the necessary thermodynamics
the overall engine performance. functions in the Numerical Propulsion System Simulation
(NPSS) is described [3]. These functions are then used as the
INTRODUCTION basis for constructing a general loss analysis element in the
The fundamental thermodynamic function of all aircraft NPSS-native programming language. This element is
engines is to convert work potential stored in the chemical described in detail and it is applied to the analysis of a mixed
bonds of the fuel into useful thrust work. The efficiency with flow turbofan engine model to illustrate results obtained with
which this occurs is typically measured in terms of a system- these tools.
level metric such as specific fuel consumption. This overall
system performance is the result a thermodynamic cycle ANALYSIS OF LOSS IN ENGINE CYCLE MODELS
consisting of many dozens of flow processes that occur inside The first law of thermodynamics is a statement of
an engine, each process contributing a small portion of loss that conservation of energy. The second law of thermodynamics
is ultimately manifested as an incremental reduction in system states that the entropy of a closed system can only remain the
performance. It is logical to think that having detailed same or increase. Typical textbook presentations usually
estimates of these various losses is a necessary prerequisite to express the second law in inequality form as: S1≤S2 where S1 is
reducing them and therefore improving the system as a whole. the total system entropy at some time and S2 is total system
Unfortunately, all past and current cycle analysis codes are entropy at a later time. Another way of stating this same thing
structured such that detailed loss information is not readily is: S1 + ∆S = S2 where S1 and S2 are as before, and ∆S is the
available and can only be obtained after great effort and entropy generated in the system between states 1 and 2. It has
inconvenience to the user. been shown that the entropy generation in the system is related
This situation has been one of the biggest barriers to to the reduction in maximum work that can be done by the
application of loss analysis methods to engine performance system on its environment: Wlost = Tref∆S where Wlost is the

1 Copyright © 2005 by ASME

work potential that was lost due to the generation of entropy in Inlet State
the system and Tref is the reference (dead state) temperature of (Input Work Potential)
the surrounding environment [4]. Thus, an alternate and
perhaps more intuitive expression of the second law of
thermodynamics is W1 - Wlost = W2 where W1 is the maximum Work
work that the system could theoretically do on its environment Output
at state 1 and W2 is the maximum work potentially available at Arbitrary Flow
Process Dissipation of
state 2.
Work Potential
This simple concept of work potential is the crux of the ith Component
loss analysis methods discussed in this paper [5]. The central
concept is that a fluid at a given temperature, pressure, and
composition contains a thermodynamically precise and Outlet State
(Output Work Potential)
quantifiable amount of work potential. It is the transfer and
loss of this work potential that is the quantity of fundamental Figure 1: "Black Box" Representation of Component
interest for prime-movers such as aircraft engines. Flow Processes.
For example, the conditions at a given flow station in an functions are currently available in the production NPSS
engine can be precisely quantified using standard cycle analysis system (version 1.6.1G and later).
methods. If the fluid at this flow station is at a temperature and
pressure other than the ambient temperature and pressure, we NEW FLOWSTATION MEMBER FUNCTIONS
know that it is possible to extract work from the flow passing In implementing new FlowStation member functions, a
through this flow station. The second law of thermodynamics number of conditions had to be satisfied in order for the
sets an upper bound on the work that can be extracted from this proposed modifications to be deemed acceptable for inclusion
flow. If this information is calculated for every flow station in in a production NPSS release. Among these, the work potential
an engine model, it can be used to deduce the losses inside each functions can not impact or in any way alter existing flow
engine component. This information is useful not only for station calculations; the source code modifications should be
understanding component performance, but also in minimal and should also be transparent to the user; the number
understanding the performance of the engine as a whole. of functions to be implemented should be the minimum
To gain a better understanding of how the loss analysis necessary and those implemented should be simple to use; the
works, consider a single component in isolation. One can calculations should have negligible impact on calculation
conceptualize any arbitrary component as a “black box” that speed; and the functions should not contain any internal
has one or more fluid streams entering and exiting Figure 1). In convergence loops or add any new failure modes to the
addition, there may be shaft work entering or exiting the FlowStation object. The present NPSS implementation of work
component, and possibly heat transfer through the component potential calculations satisfies all these conditions.
surfaces (we will neglect heat transfer in this discussion). In As this project is the first time this type of loss analysis
the NPSS environment, these streams are organized as four functionality has been implemented in a production cycle
types of port objects: fuel ports, fluid ports, bleed ports, and analysis code, only the simplest and most useful work potential
shaft ports [6]. All energy and work transfer occurring into or functions were implemented. These are manifested as three
out of a component must pass through one of these ports. Since new FlowStation member functions: getExergy,
NPSS uses 1-D flow station representations of the engine getIdealWork, and getIdealEnthalpy. The first two
system, it is assumed that each port can be represented by its are used principally for loss analysis while the third is useful
average station properties. For airbreathing propulsion for efficiency calculations in various compression and
applications, the thermodynamic state of the fluid is specified expansion elements.
by total temperature (Tt), total pressure (Pt), and fuel-air ratio
(FAR).* This, in addition to total mass flow rate (W), is getExergy Flow Station Function
sufficient to describe total energy and work potential flux The getExergy FlowStation member function has the form:
passing through a fluid or bleed port. In the case of a shaft, the real getExergy (real Tref, real Pref);
energy and work transfer are equal and are given by the shaft where Tref and Pref are the temperature and pressure used as
power. Let us ignore work potential flux in fuel ports for the the dead state point of reference for work potential calculations.
present time. The total loss inside this component can be Tref and Pref are typically taken to be ambient temperature and
calculated by merely summing the net input and output streams pressure for most analysis purposes. getExergy returns the
of work potential. This procedure is repeated for every model
mass-specific exergy content of the mass flux passing through
element to yield total losses throughout the engine.
that flow station. It is evaluated in NPSS as:
The loss analysis method described above was
implemented by first extending the FlowStation object to exergy = ht (Tt , Pt , FAR) − ht (Tt ,ref , Pt ,ref , FAR) −
(1)
include new member functions for calculating work potential, Tref [ s (Tt , P, FAR) − s(Tt ,ref , Pt ,ref , FAR)]
then implementing the port summing function in a separate
analysis element. Since the NPSS FlowStation object is part of where subscript ‘t’ denotes stagnation conditions, ‘h’ is mass-
the code base, implementation of the desired functions required specific stagnation enthalpy, ‘s’ is mass-specific entropy, Tt, Pt,
consent from the NPSS development community. These FAR define the state at the flow station of interest, and the
reference state is given by Tt,ref, Pt,ref, and FAR.
* More precisely, thermodynamic state is specified by any two intensive
thermodynamic parameters, relative velocity, and FAR (which determines
composition of the fluid).

2 Copyright © 2005 by ASME

 Carnot Work = exergy = (h-href)-Tref(s-sref) getIdealWork(Pref): move
along S=const. line until P=Pref
s - sref

Temperature
FlowStation P i
@ Tt, Pt P=
exergy=

T - Tref
carnotWork=

Enthalpy
getExergy(Tref,Pref) FlowStation: hi, Pi
getIdealWork(Pref)
Tref measures this distance
Dead State @ Tref, Pref P=P ref

Specific Entropy
sref

Entropy
Figure 2: Geometric Interpretation of Exergy Expressed in Figure 4: Definition of the getIdealWork Function
a Temperature-Entropy Diagram. Depicted on a Mollier Diagram.
Exergy physically corresponds to the work that would be kinetic energy as part of the exergy calculation. Appendix A
obtained if one were to put a Carnot device between the provides further discussion on these points.
FlowStation conditions and an infinite reservoir at Tref, Pref Figure 3 is an exergy contour plot showing mass-specific
conditions (Figure 2). The value returned from getExergy is exergy of dry air as a function of total pressure and temperature
therefore equal to the mass specific work that would be assuming a fuel-air ratio of 0.0 and reference
produced by such a Carnot device operating between these two temperature/pressure of 518R/14.7 psia, respectively. This plot
conditions. It represents an upper limit on the work that can be was created by evaluating the getExergy function over a
extracted from a flow when taking it from Tt, Pt into matrix of temperatures and pressures (at constant reference
equilibrium with its environment at Tref, Pref. conditions). The range of temperatures and pressures shown on
It should be noted that the most general definition of this plot are intended to be representative of those encountered
exergy includes many components of work potential beyond in modern aircraft gas turbine engines. The exergy content of
temperature and pressure equilibrium with the environment. air at the temperatures and pressures typical of the conditions at
For example, chemical equilibrium, species concentration, station 4 is quite appreciable. It is, in fact, significantly higher
voltage potential, gravitational potential, electromagnetic than the amount of work realized from typical turbine
radiation, nuclear, and many other components all contribute to expansion processes used today, suggesting that there is far
the total exergy of a substance. However, the vast majority of more exergy work potential available in turbine inlet flows than
these additional exergy components are of little interest for is actually extracted in practice.
cycle analysis applications and are ignored in the getExergy The source of this apparent discrepancy is the assumption
function calculations. Furthermore, getExergy assumes a of equilibrium temperature and pressure enforced by the
constant gas composition (i.e. FAR) and therefore does not definition of exergy. While it is relatively common to expand a
include exergy available due to differences in species partial high enthalpy flow to ambient pressure in an aircraft gas
pressures between flow station and reference conditions. turbine, the attainment of equilibrium temperature implicitly
Finally, getExergy calculations use the flow station total requires the inclusion of some means for exhaust flow heat
temperature and pressure, thereby implicitly including gas transfer in order to reach the reference temperature. The weight
4000 4000
800
650
3600 3600
700 600

3200 3200 550

600
500
2800 2800
500
Temperature (R)

Temperature (R)

450

2400 2400 400

400
350
350
2000 2000
300 300
1600 250 1600 250

1200 200 1200 200

150
800 Gas Tab Thermo Package 150 800
10 Dry Air, 1 atm Reference Press.
0 100
50

50 Gas Tab Thermo Package

518 R Reference Temperature Dry Air, 1 atm Reference Press.
400 400
10 20 30 40 50 10 20 30 40 50
Pressure (atm) Pressure (atm)
Figure 3: Contours of Flow-Specific Exergy as a Function Figure 5: Contours of Gas Specific Power as a Function of
of Temperature and Pressure (BTU/lbm). Temperature and Pressure (BTU/lbm).

3 Copyright © 2005 by ASME

and volume constraints placed on aircraft engines typically getIdealEnthalpy(Sref): move
make it impractical to recover waste heat to improve cycle along P=const. line until S=Sref
efficiency. This is a serious limitation on the usefulness of
exergy as a measure of useable flow work potential in aircraft Pi
engines and is the reason for the inclusion of the P=
getIdealWork function as an alternative method for
FlowStation: hi, Pi

Enthalpy
evaluating flow work potential.

getIdealWork Flow Station Function getIdealEnthalpy(Sref)

The getIdealWork FlowStation member function has measures this distance
the form:
real getIdealWork (real Pref); Entropy

Sref
where Pref is the pressure used as the dead state point of
reference for work potential calculations. Pref is typically Figure 6: Definition of the getIdealEnthalpy Function
taken to be ambient pressure for most analysis purposes. Depicted on a Mollier Diagram.
getIdealWork returns the mass-specific ideal compression pressure—which is the intersection of the P=Pt and S=Sref
or expansion work required to get from the current flow station lines. The definition of getIdealEnthalpy is illustrated in Figure
pressure to the user-specified pressure, Pref, while moving 6.
along a constant entropy line (Figure 4). In other words, The getIdealEnthalpy function is a useful utility for
getIdealWork returns the mass specific work that would be evaluating component efficiency. For example, compressor
produced if the flow at the flow station conditions were passed efficiency in the compressor element would be evaluated as:
through an ideal turbine (or compressor) to expand (compress) eff=(Fl_0.getIdealEnthalpy(Fl_I.S)-
it from Pt to Pref. The exhaust temperature corresponding to Fl_I.ht)/(Fl_0.ht-Fl_I.ht)
this imaginary expansion process is a fall-out from the This function eliminates the need to instantiate auxiliary
calculations. The calculation is implemented in NPSS as: FlowStation objects in the compressor element (or map
idealWork = ht ( Pt , Tt , FAR) − ht ( Pt ,ref , s = const., FAR) (2) subelement) to do calculations for “ideal” compression
where ‘s=const’ implies a constant entropy process. Thus, conditions.† Finally, take note of the difference between the
getIdealWork can be viewed as a specialized form of the getIdealWork and getIdealEnthalpy functions:
getExergy function wherein only pressure equilibrium is getIdealWork returns the vertical distance between two
enforced when calculating the work potential. constant pressure lines on an H-S diagram while
The concept of an imaginary expansion or compression as getIdealEnthalpy returns the distance between the H=0
a measure of work potential is not new and has been used for axis and a constant pressure line. One is a delta-enthalpy while
many years, though in limited ways. For example, the power the other is an absolute enthalpy.
output of core engines is typically quoted in terms of “gas
horsepower,” which is the same as getIdealWork. Other MODEL LOSS ANALYSIS ELEMENT
names for this quantity are “gas specific power,” and “available The implementation of the getIdealWork and
energy.” getExergy functions in the FlowStation object greatly
Results from the getIdealWork function are shown in simplifies the calculation of thermodynamic losses in engine
Figure 5. This is a contour plot of gas specific power as a components. As mentioned previously, the components can be
function of temperature and pressure using air as the working viewed as “black boxes” and the loss in the component
fluid and assuming the reference pressure is sea-level ambient. calculated based solely on fluid and shaft port conditions
The plot was generated by evaluating the getIdealWork entering and leaving the component. Returning to the earlier
function over a matrix of temperature/pressure combinations. compressor example, the calculation of loss in the compressor
Note that the magnitude of work potential is relatively close to element (assuming no bleed) would be given by:
that actually extracted using modern turbomachinery and that compressorLoss=(Fl_O.ht-Fl_I.ht)-
the shape of the contours is different from that of the exergy Fl_O.getIdealWork(Fl_I.Pt);
plot. This basic calculation procedure is easily generalized and
automated such that every element in an engine model can be
getIdealEnthalpy Flow Station Function analyzed quickly and easily. The logical approach for doing
The getIdealEnthalpy FlowStation member function this is to create a special loss analysis element that can be
has the form: appended to the model in much the same way as the
real getIdealEnthalpy (real Sref); “EngPerf” (engine performance summary) element is
where Sref is a reference entropy level. The currently used. This was the genesis for the development of the
getIdealEnthaply function is similar to the getIdealWork NetFlux analysis element.
function except that instead of moving along a constant-entropy The objective of NetFlux is to give NPSS users the
line to reach Pref, getIdealEnthaply moves along a capability to accurately analyze losses in any NPSS model with
constant pressure line (P=Pt) to reach Sref. In other words, minimal effort. The only effort required to get loss analysis
getIdealEnthalpy returns the mass-specific enthalpy that results is the instantiation of the NetFlux element and the
would have resulted from ideal compression (expansion) along † Incidentally, the same calculation can be done using getIdealWork:
the S=Sref line until reaching the current flow station eff=Fl_O.getIdealWork(Fl_I.Pt)/(Fl_0.ht-Fl_I.ht)

4 Copyright © 2005 by ASME

definition of the thermo quantity to be conserved. NetFlux using a FlowStation member function. Most will produce
can also be used for checking conservation of mass, nonsensical results, with getht(), getimp(),
momentum, or energy in a model and readily reveals any areas getIdealWork(), and getExergy() being the main
where the model is failing to satisfy conservation equations. FlowStation functions of interest.
There are three main parts to the NetFlux element. The first
is the PortList object, which is a generic data structure to MIXED FLOW TURBOFAN ANALYSIS EXAMPLE
contain all input and output flux quantities through all element A generic mixed flow turbofan (MFTF) is used as a case
ports. NetFlux creates one PortList object for each study to demonstrate typical results available from the NPSS
element in the model. These objects are instantiated inside loss the analysis functions and elements described herein. This
NetFlux and are named according to the name of the top- model is implemented in the NPSS-native programming
level element they represent. language and is representative of current low bypass MFTF
There are two member functions in NetFlux: engines in terms of model fidelity and engine performance
verify() and calculate(). The verify function levels. A schematic diagram of this model is shown in Figure
instantiates all PortList objects, one for each element in the 7. The element names are denoted in capital letters while the
model. Because the PortLists are instantiated in the verify connecting flow stations are denoted with numbers. Figure 7
also shows that this model features two spools, an afterburner,
function, NetFlux should be instantiated only after all
and several cooling flow circuits.
elements containing ports are instantiated. The calculate
This model was modified to perform loss calculations with
function simply loops through all ports in each element and
the addition of a few lines of NPSS code:
sums the total flux entering and exiting each type of port (fluid,
Element NetFlux MASS {
fuel, and shaft). The NetFlux element currently supports all
fluxType="";
types of ports except thermal port. It then stores these values in switchMassSpecific="True";
the proper PortList object and sums the total flux in all switchActive="On";
ports to obtain net generation/destruction of the flux quantity. }
While NetFlux was created to analyze model losses, it is solverSequence.remove("MASS");
also a general utility element for checking conservation of any postsolverSequence.append("MASS");
thermo property in an engine model. For example, the default
This code acts to instantiate a new element of type NetFlux
settings for the NetFlux element will cause it to sum the total
called “MASS” which is used to check model conservation of
mass flow into and out of each element fluid/bleed port. It will
mass (the default is fluxType=“”, which causes the element
then calculate the net mass flux generated or destroyed in each
to conserve mass). This MFTF example will demonstrate
model element, which presumably should be zero for a steady-
application of the NetFlux element to check conservation of
state model. In the case of transient models, this element
mass and energy as well as calculate loss in exergy and ideal
calculates the rate of change of fluid storage in all elements at
work. All cases are analyzed at sea level static conditions and
each time step. One can check conservation of energy in an
intermediate rated power setting. Results can just as easily be
engine model by simply specifying total enthalpy as the input
obtained at any other flight condition or power setting.
for the NetFlux element. It will automatically multiply by
total flow rate, account for shaft ports, etc. in order to yield
Conservation of Mass
calculations for net energy input/output from each element.
To start, let us use NetFlux to check conservation of
One NetFlux element must be instantiated for each
mass in this model. Results from this analysis are shown in the
thermo quantity to be conserved in the model. In general
first column of Table 1 while the corresponding FlowStation
NetFlux can track any thermo property that can be accessed
mass flows are shown in the first column of Table 2. This table
Primary Flowpath 14 15 18
Shaft DCT15 LINER_IN 97
Cooling Flow Circuit DCT14 AB Liner &
Flow Station
(Bypass Duct) Nozzle Cooling
Flow Circuit
Flow Resistance
LOSHAFT 16
Compressor/Turbine
HISHAFT ACCY

Inlet/Nozzle
FUEL36 FUEL6
LINER_OUT

FLOWEND9

0 2 21A 21 25 3 4 41 42 48 49 5 56 58 6 65 7 9
DCT21

DCT58
SPL21

DCT5

AMB CMP25
BRN36 MIXER AB6
INL1 CMP2 TRB41 TRB48 NOZ9
G3W41
(Fan) & G3W42
BLD41: HPT Non-Ch
G29W42
BLD42: HPT Chargeable Cooling
G26W5
FLcustBleed & G27W5
OVRBRD BLD5: LPT Non-Chargeable Cooling

Figure 7: Schematic Diagram of Mixed Flow Turbofan Cycle Model.

5 Copyright © 2005 by ASME

shows the net flux for each MFTF model element, where net energy in the burner BRN36. This is due to the way incomplete
flux out of an element is defined as positive. The element combustion is handled in the burner element calculations—a
names in the first column of Table 1 and the flow station portion of the fuel’s lower heating value (LHV) disappears due
numbers in the first column of Table 2 correspond to those to incomplete combustion. As one might expect, the AMB and
defined in Figure 7. Note that the FlowStart element FLOWEND9 elements have large net creation and destruction
“AMB” has a net creation of mass flow while the FlowEnd of energy, respectively. The FuelStart element FUEL36
has a net destruction. Furthermore, 3.5 lbm/s of fuel mass is also has a large positive energy flux due to the LHV of the fuel
injected in the combustor while 0.8 lbm/s flow is extracted for flow it contains. Summing all energy sources and sinks in the
customer use. Summing the net mass flux in/out of each model shows that there is 21.5 BTU/s more energy leaving the
element, the total across all elements is shown to be zero model than entering. This error is small relative to the total
implying that this model accurately conserves mass flow. energy flux in the model and is due to residual errors remaining
after the cycle balance calculations. This error can be driven
Conservation of Energy closer to zero by simply tightening the solver tolerances.
The NetFlux element can be used to provide an
independent check of conservation of energy by setting the Gas Specific Power Losses
fluxType variable to sum net enthalpy into/out of each NetFlux can be used to calculate available energy (gas
element: horsepower) losses throughout the model by setting the
fluxType=”ht”; fluxType variable to:
The analysis results are shown in the second column of Table 1, fluxType=”getIdealWork(AMB.Ps)”;
where the components named on each row correspond to the The results of this analysis are shown in the third column of
model components shown in Figure 7. Note that the net energy Table 1. For example, these results show that the pressure drop
flux through the model is not zero due to the way the engine is due to the afterburner flameholders causes a loss of 478.6
modeled. First, the model has a net energy output of 15.1 BTU/s (677 HP), an appreciable penalty to overall
BTU/s into the accessories section. From the model’s point of performance. The accessories yield a loss of 15.1 BTU/s, as
view, the accessories appear as an energy sink. Likewise, the before. Although there is a large energy flux in the AMB
high and low spool windage losses appear as a net destruction element, there is no appreciable work loss.
of energy in the model. This energy is implicitly removed by Closer scrutiny of the ideal work losses listed in Table 1
the oil system, the details of which are not modeled here. The reveals some surprising and counterintuitive results. First, the
278.8 BTU/s used for customer bleed appears as a net energy BRN36 element appears to have a large loss of gas specific
sink for the same reasons. Note also the apparent large loss in power. This is a manifestation of the bookkeeping system

Table 1: Net Creation/Destruction of Mass, Energy, Ideal Table 2: Flow Station Conserved Quantities for SLS MFTF
Work, and Exergy in Each Model Element. Example Case.
Component Mass Energy IdealWork Exergy Flow Station Mass Energy IdealWork Exergy
lbm/s BTU/s BTU/s BTU/s lbm/s BTU/pps BTU/pps BTU/pps
AMB 245.2 30450.9 -0.2 0.0 Station 0 245 124 -6 -6
INL1 0.0 0.0 0.0 0.0 Station 2 245 124 -6 -6
CMP2 0.0 0.0 -1219.4 -1207.4 Station 21A 245 183 54 54
SPL21 0.0 0.0 0.0 0.0 Station 21 163 183 54 54
DCT21 0.0 0.0 -60.6 -58.6 Station 25 163 183 53 53
CMP25 0.0 0.0 -2183.2 -1977.0 Station 3 122 374 232 233
Primary Flowpath

FUEL36 3.5 65068.8 65068.8 65068.8 Station 4 125 875 509 613
BRN36 0.0 -780.8 -29500.0 -16500.0 Station 41 146 802 470 553
Primary Flowpath

BLD41 0.0 0.0 277.8 -641.7 Station 42 160 573 233 321
TRB41 0.0 0.0 -2893.9 -1099.3 Station 42 146 591 239 335
BLD42 0.0 0.0 -916.1 -800.2 Station 49 166 476 133 223
BLD5 0.0 0.0 -503.4 -376.8 Station 5 160 482 135 228
TRB48 0.0 0.0 -1196.8 -450.6 Station 56 166 476 130 222
DCT5 0.0 0.0 -481.4 -180.6 Station 58 239 387 107 161
MIXER 0.0 0.0 233.9 -2332.0 Station 6 239 387 105 160
DCT58 0.0 0.0 -494.4 -224.8 Station 65 239 387 103 159
FUEL6 0.0 0.0 0.0 0.0 Station 7 247 380 102 155
AB6 0.0 0.0 -478.6 -216.0 Station 9 247 380 102 155
LINER_OUT 0.0 0.0 3.9 -219.1 Station 14 81 183 54 54
Flowpath
Bypass

NOZ9 0.0 0.0 0.0 0.0 Station 15 81 183 53 53

FLOWEND9 -247.9 -94400.0 -25300.0 -38400.0 Station 16 73 183 52 52
Flowpath

DCT14 0.0 0.0 -60.5 -58.4 Station 18 81 183 52 52

Bypass

DCT15 0.0 0.0 -60.9 -58.4 G26W5 1 330 174 178

LINER_IN 0.0 0.0 0.0 0.0 G27W5 4 330 174 178
Flow Ckts
Cooling

ACCY 0.0 -15.1 -15.1 -15.1 G29W42 1 374 232 233

HISHAFT 0.0 -60.1 -60.1 -60.1 G3W41 21 374 232 233
LOSHAFT 0.0 -6.4 -6.4 -6.4 G3W42 11 374 232 233
OVRBRD -0.8 -278.8 -153.8 -156.1 Station 97 8 183 52 52
TOTAL 0.0 -21.5 -0.2 30.2 FlcustBleed 8 340 187 190

6 Copyright © 2005 by ASME

chosen for the NetFlux element. The present implementation modeled using a nozzle thrust coefficient, so it was puzzling
assumes that the entire lower heating value of incoming fuel in that the NetFlux element returned zero loss for this element.
fuel ports is available as ideal work (and exergy). Of course, Upon further inspection, it was found that the nozzle element
only a fraction of the fuel’s heating value is realized as correctly applies a thrust decrement due to thrust coefficient,
available energy after combustion. As a result, the combustor’s but this is never translated into a reduced stagnation pressure in
fuel input port appears to have a large ideal work input and a the nozzle exit flow station. This is a flaw in the nozzle
much smaller ideal work output at the outlet with a net loss to element calculations, revealed by analysis results from the
match. The available energy input due to combustion is NetFlux element. The NetFlux element is very useful for
confounded with available energy loss due to combustor finding these types of errors in model calculations.
pressure drop and incomplete combustion. The analysis
methods described in Ref. 2 are required if one desires to Exergy Losses
estimate these individual loss contributions. To calculate exergy losses in the model, the fluxType
A second counterintuitive result is the apparent generation variable should be set to:
of available energy in the bleed element BLD41 and in the fluxType=”getExergy(AMB.Ts,AMB.Ps)”;
MIXER element. The generation of available energy in a The exergy losses are shown in the last column of Table 1 with
component appears at first glance as though it should violate the corresponding flow station exergies shown in the last
one or more laws of thermodynamics. In fact, it does not. column of Table 2. The flow exergy always decreases in all
Rather, it is a manifestation of the definition of available components, including those components where two streams
energy. It can be shown that ideal work will, in general, be are mixed. Note also from Table 2 that the flow exergy is
created whenever a high and a low temperature stream are always higher than the corresponding ideal work (available
mixed. This stands in contrast to exergy where the mixing of energy). This is consistent with the intuitive expectation that
hot and cold streams always results in destruction of exergy. ideal work should be less since it is defined as work available
The explanation for this apparent generation of available through pressure equilibrium only whereas exergy includes
energy lies in the fact that the available energy of the “warm” pressure and temperature equilibrium with the environment.
mixed stream is higher than the average of the “hot” and “cold” All shaft power losses are the same in all three cases of
unmixed streams. This is depicted in the Mollier diagram of energy, ideal work, and exergy. Shaft work, flow velocity, and
Figure 8. This figure shows two constant pressure lines, one at potential energy (height) are all forms of energy with perfect
ambient pressure and the other at a higher pressure. The order. All of the energy stored in any of these forms is
vertical distance between the contours is the available energy. therefore available to do work. As a result, the definitions for
By conservation of energy, the enthalpy of the mixed stream exergy, ideal work, and energy all reduce to identical forms for
(assuming equal mass flow rates and pressures) is simply the these modes of work transfer.
average of the enthalpies of the hot and cold streams. This It should be obvious from this example that the
corresponds to an entropy slightly higher than the mass average fluxType variable can be set to any character string which,
entropy of the incoming streams. As a result the ‘∆h’ of the when evaluated by FlowStation, will return a real value. For
mixed flow is greater than the average of the ∆h of the two example, one could check model conservation of momentum by
separate streams, resulting in an apparent increase in available specifying the impulse function as the fluxType to be
energy. The vertical ∆h distances are compared at left in the conserved.
figure—note that 2∆hmix>∆hhot+∆hcold. Thus, the generation of
available energy is purely a thermodynamic gas property FUTURE WORK
phenomena and occurs because the lines of constant pressure The proof-of-concept implementation described in this
on a Mollier diagram both curve and diverge. Proper paper is the basic infrastructure needed for fast and accurate
calculation of loss in the mixer and bleed elements requires an loss analysis of engine cycle models. Because it is
additional correction to account for this phenomena. implemented in an NPSS-native environment, it provides a
A final curious result shown in Table 1 is the nozzle loss good starting point for further developments needed to
(or lack thereof). The calculated nozzle loss is zero for both the “productionize” the analysis tool. It is hoped that this tool or
ideal work and exergy calculations. In fact, nozzle losses are something similar can become a part of the standard NPSS
distribution at some point in the future.
There are several areas still needing further development in
hhot order to make the tool truly “production ready.” First, the
current implementation ignores exergy contained in fuel ports.
(hcold+hhot)/2
A more thorough handling of fuel exergy that accounts for fuel
Enthalpy, h

=hmix
composition, temperature, and pressure is desirable. This is
∆hhot especially relevant to proper loss analysis of SCRAM and
rocket propulsion systems (wherein large heat fluxes are
hcold ∆hmixed absorbed by the fuel). Also, additional logic in the NetFlux
∆hcold
Pamb element to handle thermal ports will be required for accurate
analysis of these systems.
A more general treatment of ideal work for situations
Entropy, s involving mixing of unlike streams is also needed. At present,
the user must manually calculate theoretical “ideal work
Figure 8: "Creation" of Available Energy Via Mixing of generation” due to mixing of streams. An automated means for
Cold and Hot Streams.

7 Copyright © 2005 by ASME

doing this would be preferred. Finally, the use of NetFlux 4

Partial Pressure Exergy (Btu/lbm)

for checking conservation of momentum via the impulse
function is not well-validated and needs further work. 3.5
3
CONCLUSIONS 2.5
Work potential analysis uses absolute loss as a measure of
2
performance instead of examining and comparing relative
efficiencies as is conventionally done. This allows developers 1.5
to directly compare the losses within the components of a 1
system against each other, as well as compare overall losses of JP-8/Air
0.5
any two systems regardless of their architecture. Additionally, GasTab Thermo Package
it is simple to determine how much of the losses within 0
components and systems are inherent to the thermodynamic 0 0.02 0.04 0.06 0.08 0.1
cycle, and how much of that loss can eventually be eliminated Fuel-Air Ratio
through component improvements. Attention may then be Figure 9: Difference in Calculated Exergy (Vitiated Versus
focused not on areas of the propulsion system where the Non-vitiated Dead State Assumptions).
greatest losses exist, but on the areas where the greatest room Even after temperature and pressure equilibrium with the
for reduction of loss exists. This type of knowledge is essential environment are reached, there is still a small component of
to making well-informed tradeoffs between design alternatives. exergy available. This is due to the fact that the chemical
This paper described and demonstrated a new general loss composition of the working substance will not, in general, be in
analysis tool that directly facilitates the above objectives. This equilibrium with its environment. Specifically, the
tool consists of three new FlowStation member functions and a environment is non-vitiated and therefore any substance with a
new element added to the NPSS system. This tool is trivially fuel-air-ratio other than 0.0 will contain a component of exergy
simple to use and can currently provide loss results measured due to the differences in gas composition. Imagine a third step
relative to available energy or exergy ideal. The system is using a semi-permeable membrane to get additional PdV work
extensible so that additional figures of merit can be by equilibrating partial pressures with the environment.
implemented with minimal difficulty. Finally, this analysis tool Equilibrium is reached only when the partial pressures of all
has the added side-benefit of providing a simple and easy constituents are equal to the environment (equal partial
means for summarizing and checking conservation of mass, pressures of CO2, H2O, etc.).
momentum, and energy in any arbitrary cycle model. This third process is equivalent to defining the dead state
as having FAR=0. Therefore, the chemical component of
ACKNOWLEDGMENTS exergy is a function of FAR only, as shown in Figure 9. Note
We would like to thank NASA Glenn Research Center for that its contribution to exergy is relatively small in comparison
supporting elements of this work under contract NAS 3-03115. to the other two and is only a significant consideration at
In addition, we would like to thank Mr. Tom Lavelle at NASA temperatures and pressures close to the dead state.
and Mr. Mike Lucas at GE Aircraft Engines for their assistance. The exergy available by equilibration of chemical
composition is of little interest in virtually all practical cycle
APPENDIX A—EXERGY AND ITS RELATION TO analysis applications. Therefore, the getExergy() function
DEFINITION OF DEAD STATE implemented in NPSS defines the dead state only in terms of Tt,
There are some subtleties associated with the definition of and Pt (leaving FAR to float at current conditions). This is
the exergy dead state that one should be aware of before using equivalent to deleting the partial pressure component of exergy.
it as an analysis tool. By definition, exergy is the total work One can easily include it a posteriori by applying a small delta
that could be extracted from a substance in bringing it into total as a function of FAR.
equilibrium with its environment. We have implicitly assumed
that the environment is essentially infinite (or is large enough REFERENCES
that it is not appreciably changed by extracting work from the [1] Roth, B.A., Mavris, D.N., "A Comparison of Thermodynamic Loss
substance). This general definition of exergy includes all Models Suitable for Gas Turbine Propulsion," J. of Propulsion and
possible means of work production including mechanical, Power, Vol. 17, No. 2, pp. 324-332, Mar-Apr 2001.
temperature, chemical, electrical, magnetic, nuclear, etc. Most [2] Roth, B.A., "A Work Potential Perspective of Engine Component
of these modes are of little interest for gas turbine applications, Performance," J. of Propulsion and Power, Vol. 18, No. 6, pp. 1183-
and are ignored. The ones of interest here are: kinetic energy, 1190, Nov-Dec 2002.
internal energy, and chemical composition. [3] Vivek Sanghi, B. K. Lakshmanan, V. Sundararajan, “Survey of
The kinetic energy component of exergy is accounted for Advancements in Jet-Engine Thermodynamic Simulation,” J. of
by doing all the work potential calculations using stagnation Propulsion and Power, 2000 Vol.16 No.5 Pp. 797-807.
[4] Bejan, A., Advanced Engineering Thermodynamics, 2nd Ed., Wiley,
properties. The internal energy component of exergy is 1997.
manifested as the work available by taking the substance to [5] Roth, B.A., "The Role of Thermodynamic Work Potential in
pressure and temperature equilibrium. One could imagine that Aerospace Vehicle Design," Cleveland, 16th ISABE, Sept. 2003.
this is achieved in a two-step process: first using an isentropic [6] Evans, A.L., et al., “Numerical Propulsion System Simulation's
turbine to reach pressure equilibrium (this is what National Cycle Program,” AIAA Paper 98-3113, 1998.
getIdealWork does) followed by a Carnot engine to reach
temperature equilibrium.

8 Copyright © 2005 by ASME

View publication stats